![](images/null.gif) |
![](images/null.gif) |
![](images/null.gif) |
|
EuroSIDA Analysis Suggests Much Lower PSA Cutoff for HIV-Positive Men
|
|
|
15th European AIDS Conference, October 21-24, 2015, Barcelona
Mark Mascolini
Prostate-specific antigen (PSA) above 1.5--rather than the standard cutoff of above 4--proved the best predictor of prostate cancer in a European analysis [1]. The 1.5 cutoff had good sensitivity and specificity for HIV-positive men younger than 50 or 50 and older.
EuroSIDA researchers who conducted this study noted that prostate cancer will probably become more prevalent in men with HIV as more of them live to an older age. In a 30-year Multicenter AIDS Cohort Study analysis of more than 7000 men with or without HIV infection, prostate cancer had the highest overall incidence among non-AIDS cancers, accounting for 21% of non-AIDS malignancies [2].
Although PSA testing has become routine in older men throughout the world, the EuroSIDA team observed that only limited data addresses PSA variations and practices in men with HIV and that no clear guidelines on PSA testing for HIV-positive men have emerged. In the general population, a PSA above 4 ng/mL has become the standard cutoff indicating high prostate cancer risk. Health authorities began to question that cutoff, however, when research showed prostate cancer developing in some men with a PSA under 4, while many men with a PSA above 4 did not get prostate cancer [3].
The two-part EuroSIDA study assessed variations in PSA testing among HIV-positive men across Europe and evaluated PSA cutoffs as cancer predictors. The analysis involved HIV-positive men seen at centers where at least 5% of men get a PSA test every year. No men had prostate cancer at baseline, which was the first visit or January 1, 2008, whichever came later. Follow-up continued until prostate cancer diagnosis, the last visit, or death.
Of the 4482 HIV-positive men seen at these clinics, 1318 (29%) had one or more PSA tests. Median age of tested men stood at 44 versus 41 in the whole group. Respective proportions of men who have sex with men (MSM) were 65% and 60% and nonwhites 6% and 8%. About 90% of all men had taken antiretroviral therapy. Median CD4 count stood at 519 in the PSA group and 510 in the overall group, and median viral load lay below 49 copies in both groups.
PSA testing rates have been stable in EuroSIDA men since 2010. Poisson regression analysis determined that men older than 50 were more than twice as likely to have a PSA test than men 36 to 40 (adjusted incidence rate ratio [aIRR] 2.37, 95% confidence interval [CI] 1.92 to 2.94). Men 41 to 50 years old were 37% more likely to have a PSA test than younger man (aIRR 1.37, 95% CI 1.10 to 1.70). Compared with MSM, all other men were about 25% less likely to have their PSA checked (aIRR 0.74, 95% CI 0.64 to 0.84).
Compared with men in northern Europe, those in east-central Europe were 7.5 times more likely to get tested (aIRR 7.49, 95% CI 6.42 to 8.73). Men in southern Europe (aIRR 1.72) or western Europe (aIRR 1.54) were also more likely than northern Europeans to get a PSA. Current smokers proved less likely to get tested than never-smokers, and HCV-positive men got tested less than men without HCV. Antiretroviral use did not affect chances of PSA testing.
The EuroSIDA team used a case-control analysis to explore the sensitivity and specificity of different PSA cutoffs in predicting cancer. (Sensitivity is the ability of a test to identify people who do have the outcome of interest; specificity is the ability of a test to classify people who do not have an outcome as negative.) Cases in this analysis were men diagnosed with prostate cancer after January 1, 2001 who had a prior plasma sample. Controls were men without prostate cancer after that date and with a prior sample. The researchers matched 21 cases to 40 controls for first and last sample date (+/- 2 years), age at first sample (+/- 10 years), CD4 count at first sample (+/- 200), and region of Europe.
Three quarters of cases and controls were MSM and 95% were taking antiretroviral therapy. Median age stood at 52 among cases and 51 among controls. Respective CD4 counts were 460 and 426, and respective viral loads 1.9 and 2.0 log (about 100 copies). Median latest PSA measured about 6 ng/mL in cases and 1 ng/mL in controls (P = 0.04).
Total PSA above 4, the standard cutoff for prostate cancer risk, had a specificity of 99% in predicting prostate cancer in these men but a sensitivity of only 38%. A PSA above 1.5 had a sensitivity of 81% and a specificity of 84%. For men 50 or older, 1.5 was the optimal cutoff for combined sensitivity (81%) and specificity (82%). For younger men, 1.4 was the optimal cutoff for combined sensitivity (86%) and specificity (94%).
The EuroSIDA team proposed that a PSA cutoff of above 1.5 should be considered for men with HIV infection. They called for clear guidelines on the role of PSA and prostate cancer screening and management in men with HIV.
References
1. Shepherd L, Borges AH, Ravn L, et al. Prostate-specific antigen (PSA) testing patterns for prostate cancer in HIV+ men. does one size fit all? 15th European AIDS Conference, October 21-24, 2015, Barcelona. Abstract PS5/5.
2. Seaberg EC, Breen E, D'Souza G, et al. Incidence of non-AIDS-defining cancers (NADCs) between 1984 and 2014 in the Multicenter AIDS Cohort Study (MACS) by HIV status and viral infection etiology. 6th International Workshop on HIV and Aging, October 5-6, 2015, Washington, DC. Abstract 12.
3. Thompson IM, Pauler DK, Goodman PJ, et al. Prevalence of prostate cancer among men with a prostate-specific antigen level < or =4.0 ng per milliliter. N Engl J Med. 2004;350:2239-2246. http://www.nejm.org/doi/full/10.1056/NEJMoa031918
-----------
Reported by Jules Levin
Testing patterns and predictive value of Prostate Specific Antigen in a European HIV -positive cohort: Does one size fit all?
![EACS1.gif](../images/102815/102815-1/EACS1.gif)
![EACS2.gif](../images/102815/102815-1/EACS2.gif)
![EACS3.gif](../images/102815/102815-1/EACS3.gif)
![EACS4.gif](../images/102815/102815-1/EACS4.gif)
![EACS5.gif](../images/102815/102815-1/EACS5.gif)
![EACS6.gif](../images/102815/102815-1/EACS6.gif)
![EACS7.gif](../images/102815/102815-1/EACS7.gif)
![EACS8.gif](../images/102815/102815-1/EACS8.gif)
![EACS9.gif](../images/102815/102815-1/EACS9.gif)
![EACS10.gif](../images/102815/102815-1/EACS10.gif)
![EACS11.gif](../images/102815/102815-1/EACS11.gif)
![EACS12.gif](../images/102815/102815-1/EACS12.gif)
![EACS13.gif](../images/102815/102815-1/EACS13.gif)
![EACS14.gif](../images/102815/102815-1/EACS14.gif)
![EACS15.gif](../images/102815/102815-1/EACS15.gif)
![EACS16.gif](../images/102815/102815-1/EACS16.gif)
![EACS17.gif](../images/102815/102815-1/EACS17.gif)
![EACS18.gif](../images/102815/102815-1/EACS18.gif)
![EACS19.gif](../images/102815/102815-1/EACS19.gif)
![EACS20.gif](../images/102815/102815-1/EACS20.gif)
![EACS21.gif](../images/102815/102815-1/EACS21.gif)
|
|
|
![](images/null.gif) |
![](images/null.gif) |
|
|